The study addresses the distribution, validity and temporal stability of diagnosis in first-admission patients, patterns of illness course over a two-year period, and the prognostic utility and diagnostic specificity of a set of clinical and psychosocial indicators, in particular premorbid functioning, substance abuse, suicidal behavior, and treatment experiences. Under the current grant, we established cooperative agreements with 35 in- and outpatient facilities in Suffolk County, Long Island. In each facility, a liaison screens all admissions in order to refer first-admission psychotic patients to our project. The design entails a baseline, 6- and 24- month comprehensive diagnostic and psychosocial assessment, as well as continual contact with participants every 3 months. A consensus diagnosis is established by project psychiatrists at baseline, and a longitudinally-derived best-estimate diagnosis is determined after the 6- and 24-month assessments. The original funding was approved for 3 years to demonstrate the feasibility of establishing the network and to determine the actual numbers of eligible patients. We now anticipate that the intake phase to recruit a sample of 600 patients will be completed by 1994. During the first 16.5 months of field work, 182 patients were administered baseline interviews. Six-month best estimate diagnoses available for the first 56 patients indicate that the most common diagnosis is bipolar disorder (27%), followed by schizophrenia (21%), psychotic depression (13%), and schizoaffective disorder (13%). In addition, 46% of the 56 patients met criteria for a lifetime substance abuse or dependence. During the renewal period, we will complete the field and analytic phases of the study. We also plan to maintain contact with the sample beyond the 24 month interview to prepare for future research.